Pollutant monitoring system and method guided by turbulence degree of flow field of clean room

Abstract

The invention discloses a pollutant monitoring system and method guided by turbulence degree of a flow field of a clean room, and aims to solve the problem that the existing monitoring technology lacks self-adaption and active tracking capability. The invention provides an innovative solution aiming at the defects that a fixed point position sensor has a blind area and a robot cannot deal with the change of a dynamic flow field during fixed inspection. The system comprises a mobile robot and a set of calculation and control system with a hierarchical data processing architecture. The remote computer is responsible for complex flow field analysis and decision making, the local industrial personal computer processes original data and executes bottom layer control, the load of the robot is effectively reduced, and efficient, stable and intelligent data transmission is achieved. The core of the method is an intelligent collaborative algorithm guided by the turbulence degree of the flow field. When the system detects that the flow field is disordered, the robot can start headwind search for rapid positioning, then seamlessly switches to disorder degree gradient tracking for accurate tracking, and finally the pollution source is locked. According to the cooperative mechanism of rough positioning and accurate tracking, the hysteresis problem of a traditional method is effectively solved, monitoring is changed from a passive mode to an active mode, and the real-time performance and the accuracy of clean room pollutant monitoring are greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of air environment health and safety, and relates to a pollutant monitoring system and method guided by the turbulence of the flow field in a clean room. Background Technology

[0002] Cleanrooms serve as core production environments for high-tech industries such as semiconductors, biopharmaceuticals, and aerospace. Their cleanliness directly impacts product quality, yield, and even safety. Accurate and real-time monitoring of contaminants within cleanrooms is crucial for maintaining their high cleanliness standards.

[0003] Currently, cleanroom contaminant monitoring primarily relies on two methods: sensor array detection and autonomous robotic inspection. However, both methods have fundamental flaws. Sensor array detection is a passive monitoring method, its core being the dense deployment of sensors at key locations to achieve area coverage. However, this type of method inherently lacks coverage capabilities for dynamic and complex environments. For example, some existing patents for fixed multi-point air quality sensors (such as application numbers: 202211539256.0; 202310254146.8) inherently fail to respond to dynamically changing pollution sources, resulting in numerous "blind spots" within the monitoring area. This allows contaminants to continue spreading undetected, causing greater harm.

[0004] While mobile robot inspections have achieved automation, their inspection strategies have significant limitations. Robots patrol strictly according to pre-set routes and schedules, completely unable to adapt to real-time environmental changes. For example, some patents (such as applications 202211298620.9; 202310818721.2; 202410519256.7) disclose schemes for robots to perform periodic inspections along preset routes, but their efficiency and coverage density are limited by the fixed path. When the flow field inside the cleanroom becomes turbulent due to equipment operation or personnel activity, the robot will still patrol along the predetermined route, unable to promptly detect and track areas of turbulence, thus missing the optimal opportunity for contaminant tracking and treatment.

[0005] This invention's monitoring method overcomes the limitations of traditional fixed-path inspections. Through the intelligent collaboration of two innovative algorithms, it transforms the robot from a simple "data carrier" into an "intelligent tracker," achieving rapid response and precise location of pollutants. When the system detects turbulence in the local flow field, the robot immediately activates the upwind search algorithm, rapidly moving upwind of the pollutant to quickly narrow the search area. Once close to a potential pollution area, the robot seamlessly switches to the turbulence gradient tracking algorithm, guiding it to move precisely along the direction of the fastest increase in flow field turbulence, ultimately pinpointing the center point of the pollution source. This collaborative mechanism of "first coarse location, then precise tracking" effectively solves the lag problem of traditional fixed inspection modes when facing sudden or dynamic pollution events, significantly improving the real-time performance and accuracy of monitoring.

[0006] Furthermore, this invention employs a hierarchical data processing architecture for the robot system, placing complex algorithmic calculations on a remote computer, thereby reducing the load on the mobile robot itself. Through multi-bus collaboration, it ensures high reliability of sensor data acquisition and real-time transmission of control commands. This design achieves efficient, stable, and intelligent data processing and transmission, providing a solid foundation for accurate decision-making throughout the monitoring system. Summary of the Invention

[0007] The technical problem this invention aims to solve is that, in cleanroom environments, existing pollutant monitoring and management methods lack adaptive and proactive tracking capabilities, failing to transform robots from simple "data carriers" into "intelligent trackers." Specifically, current mainstream monitoring methods, whether relying on fixed-point sensors or blindly inspecting along fixed paths, cannot immediately detect and track areas of turbulent flow. This passive monitoring not only misses the optimal opportunity for pollutant tracking and treatment but also fails to proactively and intelligently identify and prioritize potential pollutant accumulation areas, resulting in low detection efficiency, resource waste, and difficulty in effectively managing sudden local anomalies.

[0008] This invention aims to solve the aforementioned problems by providing a cleanroom contaminant monitoring system and method that, based on fixed-path robotic inspection, can adapt to changes in the flow field and guide the robot to actively track potential contaminant sources. This method utilizes sensors mounted on the robot to collect environmental parameters in real time and intelligently identifies and tracks the most unstable areas of the flow field through a flow field turbulence guidance algorithm. Once an anomaly is detected, the robot can deviate from the preset path, initiating upwind search and turbulence gradient tracking, thereby quickly and accurately locating contamination hotspots. This intelligent monitoring strategy not only significantly improves detection efficiency and response speed but also enables refined and dynamic management of the cleanroom environment.

[0009] The technical solution of the present invention will be further described in detail below through specific embodiments.

[0010] This invention provides a pollutant monitoring system guided by the turbulence of the flow field in a cleanroom. The system includes: a mobile robot for collecting environmental parameters, a purification system, and a calculation and control system.

[0011] 1. Further optionally, the mobile robot body in the system integrates an environmental parameter acquisition module and a mobility module. The environmental parameter acquisition module consists of an anemometer and a pollutant concentration sensor, used to collect real-time, high-precision flow field (wind speed, wind direction) and pollutant concentration data in the cleanroom, providing key basis for subsequent flow field analysis; the mobility module consists of a laser ranging radar and a mobile chassis, used to realize the robot's autonomous navigation and precise movement in complex cleanroom environments.

[0012] 2. Further optionally, the purification system is used to purify pollutants.

[0013] 3. Further optionally, the computing and control system consists of an industrial control computer and a remote computer. The industrial control computer receives and processes data from the environmental parameter acquisition module in real time, performs noise reduction on this data, and transmits the processed data to the remote computer; the remote computer controls the robot to move or stop by sending commands to the industrial control computer, the environmental parameter acquisition module to collect air information, and the purification system to turn on and off.

[0014] 4. Further optional, the industrial control computer serves as the local data processing and control unit of the mobile robot. Its core functions are: receiving and preprocessing the raw environmental data from the sensor module, executing preprocessing algorithms such as data noise reduction and data formatting, and transmitting the processed data and system status information to the remote computer via a wireless network; receiving control commands from the remote computer and directly controlling the movement (including walking, stopping, and turning) of the mobile robot chassis, as well as the start time of the air pump and the valve opening of the purification module.

[0015] 5. Further optionally, the remote computer serves as the central decision-making and algorithm execution unit of the entire system. Its core functions are: receiving processed data from the industrial control computer, performing high-level algorithm operations such as data storage, flow field feature analysis, flow velocity gradient calculation, and autonomous inspection algorithm; making intelligent decisions on the robot's movement direction and inspection strategy based on the algorithm results; and generating and transmitting the control commands to the industrial control computer to remotely control the robot's specific actions.

[0016] 6. Further optionally, the data transmission method between the modules of the system is as follows: the sensor module first transmits high-frequency, real-time raw data unidirectionally to the industrial control computer via the controller local area network (CAN bus); the industrial control computer transmits pre-processed data, system status, and abnormal alarms to the remote computer in real time via a Wi-Fi wireless network; the remote computer transmits high-level control commands unidirectionally to the industrial control computer via a Wi-Fi wireless network; the industrial control computer transmits high-frequency, low-level motion and operation control commands and status feedback information bidirectionally with the mobile robot chassis and the purification module via the CAN bus.

[0017] This invention provides a method for monitoring contaminants guided by the turbulence of the flow field in a cleanroom, characterized by comprising the following steps:

[0018] Step S1: Assemble the system according to claim 1;

[0019] Step S2: Initialization Settings. Based on the cleanroom site requirements, cleanliness level, and expected monitoring accuracy, set the mobile robot's inspection route and initialize various operating parameters. These parameters specifically include: wind direction stability threshold σ. stb Flow field turbulence threshold σ start Flow field turbulence convergence threshold σ stop Stagnation time τ, number of sensor data collections N, pollutant set threshold C f Maximum distance L in a single movement max Minimum travel distance L min Magnitude of turbulent steady gradient in the flow field

[0020] Step S3: Fixed-route inspection. The mobile robot performs free inspection in a preset exploration mode, characterized in that, during the movement, the robot moves a distance of L per single trip. max After moving, the robot pauses for τs, during which time the sensors collect pollutant concentration C. t Wind direction and wind speed There are N iterations in total, where t represents any point in time within the stagnation period τ, and is used to represent the moment of real-time data acquisition.

[0021] Step S4: Preliminary data processing. Using the data collected in step S3, calculate the time-averaged concentration C of pollutants at the i-th stagnation point of the robot. i and wind direction standard deviation σ i :

[0022]

[0023] Standard deviation of wind direction σ i The calculation method is as follows:

[0024] Step S4.1: Based on wind direction and speed Calculate the unit vector:

[0025]

[0026] Step S4.2: Calculate the X (east) component of the angle in the Cartesian coordinate system. and Y (north) component

[0027]

[0028] Step S4.3: Calculate the arithmetic mean of all unit vector components:

[0029]

[0030] Step S4.4: Calculate the standard deviation σ of wind direction at the i-th stagnation point. i :

[0031]

[0032] Step S5: Monitoring Strategy. Based on the time-averaged concentration C of pollutants. i With pollutant set threshold C f The relationship between magnitudes, and the standard deviation of wind direction σ i With the initiation threshold σ of the flow field turbulence start Wind direction stability threshold σ stb and the convergence threshold σ of the flow field turbulence stop The size relationship is used to initiate a monitoring strategy.

[0033] Strategy 1: C at the i-th stagnation point i Greater than C f , σ i Less than σ stb If the purification device is activated to remove pollutants, step S3 will be executed after purification is completed.

[0034] Strategy 2: C at the i-th stagnation point i Greater than C f , σ i Greater than σ stb And less than σ start If the purification device is activated to remove pollutants, step S6 will be executed after purification is completed.

[0035] Strategy 3: C at the i-th stagnation point i Less than C f , σ i Greater than σ stb And less than σ start Then proceed to step S6;

[0036] Strategy 4: C at the i-th stagnation point i Less than C f , σ i Greater than σ start And less than σ stop Then proceed to step S7;

[0037] Strategy 5: σ at the i-th stagnation point i Greater than σ stop Then proceed to step S8;

[0038] Strategy 6: In other cases, proceed directly to step S3.

[0039] Step S6: The robot initiates the headwind search algorithm, continuously moving according to the unit vector direction of the wind at its current location; and comparing the standard deviation σ of the wind direction at the current location during the movement. i With the initiation threshold σ of the flow field turbulence start and the convergence threshold σ of the flow field turbulence stop The relationship, if σ i Greater than σ start And less than σ stop Then proceed to step S7; if σ i Less than σ start Then continue with step S6; if σ i Greater than σ stop Then proceed to step S8. The specific method of the headwind search algorithm is as follows:

[0040] Step S6.1: Record the current location information X i =(x i y i ).

[0041] Step S6.2: Calculate the standard deviation σ of the wind direction at the robot's stationary point i. i And determine the step size of the headwind search algorithm:

[0042]

[0043] Where σ wind It is the wind direction stability threshold.

[0044] Step S6.3: Calculate the robot's position X at the stopping point i+1. i+1 :

[0045] X i+1 =X i +L i (-V(X i ))

[0046] Where V(X) i )=(u x,u y ) represents the wind direction unit vector.

[0047] Step S7: The robot starts the turbulence gradient tracking algorithm. The robot moves continuously according to the turbulence gradient direction at the current position, and compares the standard deviation σ of the wind direction at the current position during the movement. i Is it greater than the convergence threshold σ for flow field turbulence? stop If so, proceed to step S8; otherwise, continue with step S7. The specific method of the disorder gradient pursuit algorithm is as follows:

[0048] S7.1: Record the robot's current position X i =(x i ,y i ), and calculate the disorder gradient at the current location.

[0049] in

[0050] Δx and Δy are small spatial steps used to calculate the gradient, representing the wind volume of L in the x-direction and the component in the y-direction, respectively.

[0051] σ(x i ±Δx,y i ): at point (x i ,y i The standard deviation of wind direction after moving a small step size Δx along the positive / negative x-axis.

[0052] σ(x i ,y i ±Δy): at point (x i ,y i The standard deviation of wind direction after moving a small step size Δy along the positive / negative y-axis.

[0053] Step S7.2: Motion control, the robot updates its position to X according to the following formula. i+1 :

[0054]

[0055] a) where L i The step size factor for the current iteration, used to control the distance the robot travels along the actual direction of motion, is determined by the following formula:

[0056]

[0057] b) where υ i It is a random perturbation vector, sampled from a Gaussian distribution with a mean of zero, i.e., υ i ~N(0,ψ i 2k) where k is the identity matrix to ensure that the perturbation is independent and has the same strength in all dimensions, ψ i The formula for calculating the intensity of random disturbances is:

[0058]

[0059] In the formula ψ max ψ represents the maximum value of the random disturbance intensity. min This represents the minimum value of the random disturbance intensity, where b is the attenuation coefficient in the formula.

[0060] Step S8: The robot stops moving, determines that the stopping point is a potential contamination point, and transmits the coordinates of the point to a remote computer for recording.

[0061] Step S9: Determine the average concentration C of pollutants at potential pollution points. i Is it greater than the pollutant threshold C? f If yes, start the pollutant purification device and then proceed to step S10; if no, proceed directly to step S10.

[0062] Step S10: The robot leaves the potential contamination point, returns to its position closest to the fixed route, and executes step S3.

[0063] The beneficial effects of the above technical solution of the present invention are as follows:

[0064] This invention effectively overcomes the fundamental shortcomings of existing cleanroom monitoring methods, which lack adaptive and active tracking capabilities. Through the intelligent collaboration of two innovative algorithms, this invention achieves rapid response and precise location of contaminants. The robot first uses an upwind search algorithm to quickly locate potential areas, then seamlessly switches to a turbulence gradient tracking algorithm for precise movement, ultimately pinpointing the center of the contamination source. This collaborative mechanism of "rough location first, then precise tracking" effectively solves the lag problem of traditional fixed inspection modes, significantly improving the real-time performance and accuracy of monitoring. Furthermore, this invention adopts a hierarchical data processing architecture, placing complex calculations on a remote computer, reducing the robot's load, and ensuring efficient and stable data transmission through multi-bus collaboration, providing a solid foundation for accurate system decision-making. Attached Figure Description

[0065] Figure 1 This is a schematic diagram of the overall architecture of the system of this invention.

[0066] Figure 2 This is a schematic diagram of the monitoring method flow of an example of the present invention;

[0067] Figure 3 This is a schematic diagram of the disorder gradient tracking algorithm in an example of the present invention. Specific implementation methods

[0068] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0069] The technical problem this invention aims to solve is that existing contaminant monitoring and management methods in cleanroom environments lack adaptive and proactive tracking capabilities, failing to transform robots from simple "data carriers" into "intelligent trackers." Specifically, current mainstream monitoring methods, whether relying on fixed-point sensors or blindly inspecting along fixed paths, cannot immediately detect and track areas of turbulent flow. This passive monitoring not only misses the optimal opportunity for contaminant tracking and treatment but also fails to proactively and intelligently identify and prioritize potential contaminant accumulation areas, resulting in low detection efficiency, resource waste, and difficulty in effectively managing sudden local anomalies.

[0070] First embodiment:

[0071] This invention provides a pollutant monitoring system guided by the turbulence of the flow field in a cleanroom. The system includes: a mobile robot for collecting environmental parameters, a purification system, and a calculation and control system.

[0072] 1. Further optionally, the mobile robot body in the system integrates an environmental parameter acquisition module and a mobility module. The environmental parameter acquisition module consists of an anemometer and a pollutant concentration sensor, used to collect real-time, high-precision flow field (wind speed, wind direction) and pollutant concentration data in the cleanroom, providing key basis for subsequent flow field analysis; the mobility module consists of a laser ranging radar and a mobile chassis, used to realize the robot's autonomous navigation and precise movement in complex cleanroom environments.

[0073] 2. Further optionally, the purification system is used to purify pollutants.

[0074] 3. Further optionally, the computing and control system consists of an industrial control computer and a remote computer. The industrial control computer receives and processes data from the environmental parameter acquisition module in real time, performs noise reduction on this data, and transmits the processed data to the remote computer; the remote computer controls the robot to move or stop by sending commands to the industrial control computer, the environmental parameter acquisition module to collect air information, and the purification system to turn on and off.

[0075] 4. Further optional, the industrial control computer serves as the local data processing and control unit of the mobile robot. Its core functions are: receiving and preprocessing the raw environmental data from the sensor module, executing preprocessing algorithms such as data noise reduction and data formatting, and transmitting the processed data and system status information to the remote computer via a wireless network; receiving control commands from the remote computer and directly controlling the movement (including walking, stopping, and turning) of the mobile robot chassis, as well as the start time of the air pump and the valve opening of the purification module.

[0076] 5. Further optionally, the remote computer serves as the central decision-making and algorithm execution unit of the entire system. Its core functions are: receiving processed data from the industrial control computer, performing high-level algorithm operations such as data storage, flow field feature analysis, flow velocity gradient calculation, and autonomous inspection algorithm; making intelligent decisions on the robot's movement direction and inspection strategy based on the algorithm results; and generating and transmitting the control commands to the industrial control computer to remotely control the robot's specific actions.

[0077] 6. Further optionally, the data transmission method between the modules of the system is as follows: the sensor module first transmits high-frequency, real-time raw data unidirectionally to the industrial control computer via the controller local area network (CAN bus); the industrial control computer transmits pre-processed data, system status, and abnormal alarms to the remote computer in real time via a Wi-Fi wireless network; the remote computer transmits high-level control commands unidirectionally to the industrial control computer via a Wi-Fi wireless network; the industrial control computer transmits high-frequency, low-level motion and operation control commands and status feedback information bidirectionally with the mobile robot chassis and the purification module via the CAN bus.

[0078] Second embodiment:

[0079] This invention provides a method for monitoring contaminants guided by the turbulence of the flow field in a cleanroom, characterized by the following steps:

[0080] Step S1: Assemble the system according to claim 1;

[0081] Step S2: Initialization Settings. Based on the cleanroom site requirements, cleanliness level, and expected monitoring accuracy, set the mobile robot's inspection route and initialize various operating parameters. These parameters specifically include: wind direction stability threshold σ. stb =0.26, the threshold for initiating flow field turbulence σ start =0.35, the convergence threshold for flow field turbulence σ stop =0.6, stagnation time τ=30s, sensor data collection times N=30, pollutant set threshold C f According to relevant specifications, the maximum distance L in a single movement max=1.5m, minimum moving distance L min =0.5m, magnitude of turbulent steady-state gradient in the flow field

[0082] Step S3: Fixed-route inspection. The mobile robot performs free inspection in a preset exploration mode, characterized in that, during the movement, the robot moves a distance of L per single trip. max After moving, the robot pauses for τs, during which time the sensors collect pollutant concentration C. t Wind direction and wind speed There are N iterations in total, where t represents any point in time within the stagnation period τ, and is used to represent the moment of real-time data acquisition.

[0083] Step S4: Preliminary data processing. Using the data collected in step S3, calculate the time-averaged concentration C of pollutants at the i-th stagnation point of the robot. i and wind direction standard deviation σ i :

[0084]

[0085] Standard deviation of wind direction σ i The calculation method is as follows:

[0086] Step S4.1: Based on wind direction and speed Calculate the unit vector:

[0087]

[0088] Step S4.2: Calculate the X (east) component of the angle in the Cartesian coordinate system. and Y (north) component

[0089]

[0090] Step S4.3: Calculate the arithmetic mean of all unit vector components:

[0091]

[0092] Step S4.4: Calculate the standard deviation σ of wind direction at the i-th stagnation point. i :

[0093]

[0094] Step S5: Monitoring Strategy. Based on the time-averaged concentration C of pollutants. i With pollutant set threshold C f The relationship between magnitudes, and the standard deviation of wind direction σ i With the initiation threshold σ of the flow field turbulence startWind direction stability threshold σ stb and the convergence threshold σ of the flow field turbulence stop The size relationship is used to initiate a monitoring strategy.

[0095] Strategy 1: C at the i-th stagnation point i Greater than C f , σ i Less than σ stb If the purification device is activated to remove pollutants, step S3 will be executed after purification is completed.

[0096] Strategy 2: C at the i-th stagnation point i Greater than C f , σ i Greater than σ stb And less than σ start If the purification device is activated to remove pollutants, step S6 will be executed after purification is completed.

[0097] Strategy 3: C at the i-th stagnation point i Less than C f , σ i Greater than σ stb And less than σ start Then proceed to step S6;

[0098] Strategy 4: C at the i-th stagnation point i Less than C f , σ i Greater than σ start And less than σ stop Then proceed to step S7;

[0099] Strategy 5: σ at the i-th stagnation point i Greater than σ stop Then proceed to step S8;

[0100] Strategy 6: In other cases, proceed directly to step S3.

[0101] Step S6: The robot initiates the headwind search algorithm, continuously moving according to the unit vector direction of the wind at its current location; and comparing the standard deviation σ of the wind direction at the current location during the movement. i With the initiation threshold σ of the flow field turbulence start and the convergence threshold σ of the flow field turbulence stop The relationship, if σ i Greater than σ start And less than σ stop Then proceed to step S7; if σ i Less than σ start Then continue with step S6; if σ i Greater than σ stop Then proceed to step S8. The specific method of the headwind search algorithm is as follows:

[0102] Step S6.1: Record the current location information X i =(x i y i ).

[0103] Step S6.2: Calculate the standard deviation σ of the wind direction at the robot's stationary point i. i And determine the step size of the headwind search algorithm:

[0104]

[0105] Where σ wind It is the wind direction stability threshold.

[0106] Step S6.3: Calculate the robot's position X at the stopping point i+1. i+1 :

[0107] X i+1 =X i +L i (-V(X i ))

[0108] Where V(X) i )=(u x ,u y ) represents the wind direction unit vector.

[0109] Step S7: The robot starts the turbulence gradient tracking algorithm. The robot moves continuously according to the turbulence gradient direction at the current position, and compares the standard deviation σ of the wind direction at the current position during the movement. i Is it greater than the convergence threshold σ for flow field turbulence? stop If so, proceed to step S8; otherwise, continue with step S7. The specific method of the disorder gradient pursuit algorithm is as follows:

[0110] S7.1: Record the robot's current position X i =(x i ,y i ), and calculate the disorder gradient at the current location.

[0111] in

[0112] Δx and Δy are small spatial steps used to calculate the gradient, representing the wind volume of L in the x-direction and the component in the y-direction, respectively.

[0113] σ(x i ±Δx,y i ): at point (x i ,y iThe standard deviation of wind direction after moving a small step size Δx along the positive / negative x-axis.

[0114] σ(x i ,y i ±Δy): at point (x i ,y i The standard deviation of wind direction after moving a small step size Δy along the positive / negative y-axis.

[0115] Step S7.2: Motion control, the robot updates its position to X according to the following formula. i+1 :

[0116]

[0117] a) where L i The step size factor for the current iteration, used to control the distance the robot travels along the actual direction of motion, is determined by the following formula:

[0118]

[0119] b) where υ i It is a random perturbation vector, sampled from a Gaussian distribution with a mean of zero, i.e., υ i ~N(0,ψ i 2 k) where k is the identity matrix to ensure that the perturbation is independent and has the same strength in all dimensions, ψ i The formula for calculating the intensity of random disturbances is:

[0120]

[0121] In the formula ψ max The maximum value representing the intensity of the random disturbance is 0.05, ψ min The minimum value of the random disturbance intensity is 0.01, and b in the formula is the attenuation coefficient, which is equal to 0.05.

[0122] Step S8: The robot stops moving, determines that the stopping point is a potential contamination point, and transmits the coordinates of the point to a remote computer for recording.

[0123] Step S9: Determine the average concentration C of pollutants at potential pollution points. i Is it greater than the pollutant threshold C? f If yes, start the pollutant purification device and then proceed to step S10; if no, proceed directly to step S10.

[0124] Step S10: The robot leaves the potential contamination point, returns to its position closest to the fixed route, and executes step S3.

[0125] The beneficial effects of the above technical solution of the present invention are as follows:

[0126] This invention effectively overcomes the fundamental shortcomings of existing cleanroom monitoring methods, which lack adaptive and active tracking capabilities. Through the intelligent collaboration of two innovative algorithms, this invention achieves rapid response and precise location of contaminants. The robot first uses an upwind search algorithm to quickly locate potential areas, then seamlessly switches to a turbulence gradient tracking algorithm for precise movement, ultimately pinpointing the center of the contamination source. This collaborative mechanism of "rough location first, then precise tracking" effectively solves the lag problem of traditional fixed inspection modes, significantly improving the real-time performance and accuracy of monitoring. Furthermore, this invention adopts a hierarchical data processing architecture, placing complex calculations on a remote computer, reducing the robot's load, and ensuring efficient and stable data transmission through multi-bus collaboration, providing a solid foundation for accurate system decision-making.

Claims

1. A clean room flow field turbulence directed contamination monitoring system, characterized by, The system comprises a mobile robot for collecting environmental parameters, a purification system, and a computing and control system.

2. A cleanroom flow field turbulence directed contamination monitoring system according to claim 1, wherein, The mobile robot body in the system is integrated with an environmental parameter collection module and a moving module. The environmental parameter collection module is composed of a wind speed and direction instrument and a pollutant concentration sensor, and is used for collecting flow field (wind speed, wind direction) and pollutant concentration data in a clean room in real time and high precision, thereby providing a key basis for subsequent flow field analysis. The moving module is composed of a laser ranging radar and a moving chassis, and is used for realizing autonomous navigation and precise movement of the robot in a complex clean room environment.

3. A cleanroom flow field turbulence directed contamination monitoring system according to claim 1, wherein, The purification system is used for purifying pollutants.

4. A cleanroom flow field turbulence directed contamination monitoring system according to claim 1, wherein, The computing and control system is composed of an industrial computer and a remote computer. The industrial computer is used for receiving and processing data from the environmental parameter collection module in real time, performing noise reduction on the data, and transmitting the processed data to the remote computer. The remote computer controls the robot to move or stop by sending instructions to the industrial computer, and controls the opening and closing of the environmental parameter collection module and the purification system.

5. The computing and control system of claim 4, wherein, The industrial computer serves as a local data processing and control unit of the mobile robot, and its core functions include receiving and preprocessing original environmental data from the sensor module, performing data noise reduction and data formatting, transmitting the processed data and system state information to the remote computer through a wireless network, receiving control instructions from the remote computer, and directly controlling the movement (including walking, stopping, and turning) of the mobile robot chassis and the opening time of the purification module air pump and the valve opening degree.

6. The computing and control system of claim 4, wherein, The remote computer serves as a central decision-making and algorithm execution unit of the entire system, and its core functions include receiving the processed data from the industrial computer, performing data storage, flow field feature analysis, flow rate gradient calculation, and autonomous inspection algorithm, intelligently deciding the movement direction and inspection strategy of the robot, generating and transmitting the control instructions to the industrial computer to remotely control the specific actions of the robot. The data transmission mode between the modules of the system is as follows: the sensor module unidirectionally transmits high-frequency and real-time original data to the industrial computer through a controller area network (CAN bus); the industrial computer transmits preprocessed data, system state, and abnormal alarm to the remote computer in real time through a Wi-Fi wireless network; the remote computer unidirectionally transmits high-level control instructions to the industrial computer through a Wi-Fi wireless network; and the industrial computer transmits high-frequency and low-level movement and operation control instructions and state feedback information to the mobile robot chassis and the purification module through a CAN bus in a bidirectional manner.

7. The computing and control system of claim 4, wherein, The method comprises the following steps:

8. A clean room flow field turbulence directed contamination monitoring method, characterized by, Step S1: assembling the system of claim 1; Strategy 6: directly executing step S3 in other cases; Step S2: initialization setting, according to the clean room site requirements, cleanliness level and expected monitoring accuracy, set the inspection route of the mobile robot, and initialize various operating parameters. These parameters specifically include: wind direction stability threshold σ stb , flow field turbulence starting threshold σ start , flow field turbulence convergence threshold σ stop , stagnation time τ, sensor acquisition times N, pollutant setting threshold C f , maximum distance of single movement L max , minimum movement distance L min , flow field turbulence stable gradient size |▽σ| max ; Step S3: fixed route inspection, the mobile robot performs free inspection in a preset exploration mode, characterized in that, during movement, the robot moves a distance L each time max , after movement, the robot stops for ts, during which the sensor collects the pollutant concentration C t , the wind direction and speed N times, where t is a variable representing any one time point within the stop collection time τ, used to represent the time of real-time data collection; Step S4: Data preliminary processing, using the data collected in step S3, the robot calculates the time-averaged concentration of pollutants C at the i-th stagnation point i and the wind direction standard deviation σ i ; Step S5: Monitoring strategy, based on the time-averaged concentration C of pollutants i With pollutant set threshold C f The relationship between magnitudes, and the standard deviation of wind direction σ i With the initiation threshold σ of the flow field turbulence start Wind direction stability threshold σ stb and the convergence threshold σ of the flow field turbulence stop The size relationship triggers the monitoring strategy; Policy 1: C of the i-th stagnation point i greater than C f , σ i less than σ stb , then start the purification device to purify the pollutants, and after the purification is completed, step S3 is executed; Policy 2: C of the i-th stagnation point i greater than C f , σ i greater than σ stb and less than σ start , then start the purification device to purify the pollutants, and after the purification is completed, step S6 is executed. Policy 3: C of the i-th stall point i less than C f , σ i greater than σ stb and less than σ start then step S6 is performed; Policy 4: C of the i-th stall point i less than C f , σ i greater than σ start and less than σ stop then step S7 is performed; Policy 5: σ of the i-th stagnation point i greater than σ stop then step S8 is executed; Step S8: the robot stops moving, determines that the stopping point is a potential pollution point, and transmits the coordinates of the point to the remote computer for recording; Step S6: The robot starts the upwind search algorithm, and the robot moves continuously according to the wind direction unit vector direction of the current position; and compares the standard deviation σ of the wind direction of the current position in the moving process i with the flow field disorder starting threshold σ start and the flow field disorder convergence threshold σ stop , if σ i is greater than σ start and less than σ stop , step S7 is executed; if σ i is less than σ start , step S6 is continued to be executed; if σ i is greater than σ stop , step S8 is executed. Step S7: The robot starts the turbulence gradient tracking algorithm, and the robot keeps moving according to the turbulence gradient direction of the current position, and compares the wind direction standard deviation σ of the current position in the moving process i whether it is greater than the flow field turbulence convergence threshold σ stop ; if yes, step S8 is executed, and if no, step S7 is continued to be executed; Step S10: the robot leaves the potential pollution point, returns to a position closest to the fixed route from the position of the robot, and executes step S3. Step S9: judging whether the potential pollution point pollutant time average concentration C i is greater than the pollutant set threshold C f ; if yes, starting the pollutant purification device, and then executing step S10; if no, directly executing step S10. ​ 9. A method of cleanroom flow field turbulence directed contamination monitoring according to claim 8, wherein, The standard deviation σ of the wind direction at the i-th stagnation point in the step S4 i is calculated as follows: Step S4.1: Calculate the wind direction and wind speed from the wind vector Calculate the unit vector: Step S4.2: Calculate the X (due East) component of the angle in the Cartesian coordinate system and the Y (due North) component Step S4.3: Calculate the arithmetic mean of all unit vector components: Step S4.4: Calculate the wind direction standard deviation σ of the i-th stagnation point i : wherein wherein σ i A small value indicates that the wind direction is very stable and consistent over the measurement time; a i A large value indicates that the wind direction is highly fluctuating, turbulent and uncertain, and the wind can change direction substantially in a short time without a dominant direction.

10. The method of claim 8, wherein the method further comprises: The specific method of the inverse wind search algorithm in the step S6 is: Step S6.1: Record current position information X i = (x i y i ); Step S6.2: The robot calculates the standard deviation σ of the wind direction at the stagnation point i i and determines the step size of the upwind search algorithm: where σ wind is a wind direction stability threshold Step S6.3: The robot calculates the position X of the stagnation point i+1 i+1 : X i+1 = X i + L i (- V(X i )) where V(X i ) = (u x ,u y ) denotes the wind direction unit vector.

11. The method of claim 8, wherein the method further comprises: The specific method of the turbulence gradient tracking algorithm in the step S7 is: S7.1: Record the current position X of the robot i = (x i , y i ), and calculate the turbulence gradient Ds i at the current position wherein Δx and Δy are the small spatial steps used to calculate the gradient, representing the L component in the x and y direction respectively; σ(x i ±Δx,y i ): standard deviation of wind direction after moving a small step Δx in the positive / negative direction of the x-axis from point (x i ,y i ); σ(x i ,y i ±Δy): standard deviation of wind direction after moving a small step Δy in the positive / negative direction of the y-axis from point (x i ,y i ). Step S7.2: Motion control, the robot updates its position to X according to the following equation i+1 : a) wherein L i is the step length factor of the current iteration for controlling the distance of the robot along the actual movement direction, which is determined as b) where υ i is a random disturbance vector, sampled from a Gaussian distribution with zero mean, i.e. υ i ~ N(0, ψ i 2 k), where k is an identity matrix ensuring that the disturbances are independent and of equal strength in each dimension, ψ i is the random disturbance strength. The calculation formula is: ψ max represents the maximum value of the random disturbance intensity, ψ min represents the minimum value of the random disturbance intensity, and b is a decay coefficient in the formula.

Images